Volatility — the standard deviation of returns — quietly assumes returns follow a bell curve: symmetric, with extreme moves vanishingly rare. Real market returns do neither. They are lopsided and they have fat tails, and the two numbers that measure those departures — skewness and kurtosis — are where the risks volatility misses actually live.
Skewness: the lopsided tail
Skewness measures whether one tail of the distribution stretches further than the other. Equity returns are usually negatively skewed — a long left tail — meaning a steady drip of small gains interrupted by occasional large losses. That is the opposite of comfortable: the average looks fine, but the rare bad outcomes are worse than the rare good ones. Positive skew, by contrast, is the lottery-ticket or long-option profile: many small losses, rare big wins.
Kurtosis: the fat tails
Kurtosis measures how often extreme moves happen. A normal distribution has a kurtosis of 3; markets run well above it — they are fat-tailed. Events a bell curve calls once-in-a-century arrive every few years, because panic, leverage and herding make big moves cluster rather than scatter. Excess kurtosis (kurtosis minus 3) is reliably positive for financial returns, and it is the mathematical signature of the crashes and spikes that define real investing.
A strategy has posted small, steady gains for years with low volatility. What should skewness and kurtosis make you check before calling it low-risk?
Volatility maan leti hai returns ek clean bell curve hain — hain nahi. Skewness batati hai distribution kis taraf jhuka hai (equity aksar negative skew — chhote-chhote faayde, kabhi-kabhi bada nuksaan). Kurtosis batati hai tails kitni moti hain — market mein "100 saal mein ek baar" wale crash har kuch saal mein aate hain. Steady, low-volatility strategy (jaise option bechna) bhi negative-skew fat-tail ho sakti — dikhta smooth, phir ek crash mein saaf. Sirf volatility pe "low-risk" mat maano.
- Volatility assumes a symmetric bell curve; real returns are neither symmetric nor thin-tailed.
- Skewness measures lopsidedness — equities are usually negatively skewed (a long left tail).
- Kurtosis measures fat tails; markets far exceed the normal distribution’s value of 3.
- Normal-based risk tools (volatility, parametric VaR) understate the danger of these features.
- Steady, low-volatility returns can hide negative skew and fat tails — check the shape, not just the width.
Mark it done to track your progress through the curriculum.
Common questions
Short, direct answers to what people ask about this topic.
- what is skewness in stock returns
- Skewness measures how lopsided a distribution of returns is — whether its tail stretches further to one side than the other. Negative skew, common in equities, means the distribution has a long left tail: many small gains punctuated by occasional large losses. Positive skew is the reverse: many small losses with rare large gains, the profile of things like lottery tickets or out-of-the-money options. Because most stock and index returns are negatively skewed, the average return understates how bad the rare bad days can be, which is exactly what a symmetric volatility number fails to capture.
- what is kurtosis in stock returns
- Kurtosis measures how fat the tails of a return distribution are — how often extreme moves, far from the average, actually occur. A normal distribution has a kurtosis of 3, and anything above that is called “fat-tailed” or leptokurtic, meaning extreme events happen more often than the bell curve predicts. Financial returns are famously fat-tailed: crashes and spikes that a normal distribution would call once-in-a-century occur every few years. Excess kurtosis (kurtosis minus 3) is the usual way this is quoted, and for markets it is reliably positive.
- why are stock market returns not normally distributed
- Because markets are driven by human behaviour, feedback loops and leverage that produce far more extreme moves than a bell curve allows. Panic selling, margin calls and herding cause losses to cluster and cascade, creating fat tails (high kurtosis) and a long left tail (negative skew). The normal distribution assumes each day is independent and extreme moves are vanishingly rare, but real markets have volatility clustering — calm periods and violent ones bunch together — and crashes far larger than the model permits. Assuming normality quietly understates the risk of exactly the events that matter most.
- why does assuming a normal distribution understate risk
- Because the two ways real returns depart from normal — negative skew and fat tails — both make disasters more likely and more severe than the bell curve implies. Risk measures built on the normal assumption, like a basic volatility figure or parametric Value at Risk, therefore predict that severe losses are rarer than they truly are, giving false comfort. The 2008 crisis and other blow-ups happened partly because models treated fat-tailed markets as normally distributed. Recognising skew and kurtosis is what pushes serious risk management toward stress tests and expected shortfall.